On Thu, Nov 06, 2014 at 05:06:21PM +0100, Maxime Ripard wrote:
> Hi,
>
> On Wed, Nov 05, 2014 at 09:56:35AM -0800, Dmitry Torokhov wrote:
> > On Wed, Nov 05, 2014 at 04:07:49PM +0100, Maxime Ripard wrote:
> > > So far, the DT parsing code was only setting up the regular input axes,
> > > completely ignoring their multitouch counter parts.
> > >
> > > Fill them with the same parameters than the regular axes.
> > >
> > > Signed-off-by: Maxime Ripard <maxime.ripard@xxxxxxxxxxxxxxxxxx>
> > > ---
> > > drivers/input/touchscreen/of_touchscreen.c | 12 +++++++++---
> > > 1 file changed, 9 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/input/touchscreen/of_touchscreen.c b/drivers/input/touchscreen/of_touchscreen.c
> > > index 74d6b0eb50ac..cf2a753edd96 100644
> > > --- a/drivers/input/touchscreen/of_touchscreen.c
> > > +++ b/drivers/input/touchscreen/of_touchscreen.c
> > > @@ -42,17 +42,23 @@ void touchscreen_parse_of_params(struct input_dev *dev)
> > >
> > > maximum = of_get_optional_u32(np, "touchscreen-size-x");
> > > fuzz = of_get_optional_u32(np, "touchscreen-fuzz-x");
> > > - if (maximum || fuzz)
> > > + if (maximum || fuzz) {
> > > input_set_abs_params(dev, ABS_X, 0, maximum, fuzz, 0);
> > > + input_set_abs_params(dev, ABS_MT_POSITION_X, 0, maximum, fuzz, 0);
> >
> > Not all devices are multitouch so you shoudl not be setting multitouch
> > bits unconditionally.
>
> Hmmm, right.
>
> > In I think we should rely on driver to set capability bits properly and
> > then here test them and apply the readings. Probably also issue a
> > warning if we see max/fuzz setting but neither ABS_<N>/ABS_MT_<N>
> > capabilities.
>
> What I was aiming at was to avoid DT parsing duplication for the
> !multitouch and multitouch axis. I don't think there's a way to copy
> the parameters.
>
> The mt_init_slots might be a solution, but it does the copy the other
> way around: from the multitouch to the !multitouch axis, and without
> enabling it, which renders using both input_mt_init_slots and the
> of_touchscreen code together impossible.
>
> Is there a way to just enable an axis without calling
> input_set_abs_params? Is __set_bit enough?
Hmm, not anymore I'm afraid as you still need absinfo being allocated,
but there is input_set_capability() that is basically a fancy
__set_bit(). If you set ABS_* with it that would mean that the device
has unspecified min/max/fuzz.
And then, in DT parsing code, I'd do:
...read max and fuzz for an axis...
if (max || fuzz) {
if (!test_bit(ABS_N, dev->absbit) &&
!test_bit(ABS_MT_N, dev->absbit)) {
dev_warn("... have parameters but axis not set up...");
} else {
if (test_bit(ABS_N...)
input_set_abs_params(dev, ABS_N...);
if (test_bit(ABS_MT_N...)
input_set_abs_params(dev, ABS_MT_N...);
}
Thanks.
--
Dmitry
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
